Considering more solar and a battery

NedS

Yes, under the G98 application as submitted, they would only allow for 1kW export in total.

However, good news, just spoken with the DNO.

The limitation is due to voltage rise concerns, and potentially exceeding the 253V limit due to the number of generators already connected to our feeder from that substation.

But the DNO guy told me about a fast track application process based on SGI3 and an impedance test that is often beneficial to costumers where export limits may otherwise result. That process allows for 2 x 5kW max inverters and total export limitation of 5kW, so we are going to reapply on that basis. He said far too many installers either are not aware of this application process or do not use it (not sure if it is specific to NG, our DNO, or if it is generic)

The revised proposal would be to keep our existing 3.645kWp array on 3.0kW inverter and add the new 3.56kWp array connected to the Tesla PW3 inverter set at 5kW max (can be set at either 3.68 or 5kW) and then limit export from the property to 5kW (it doesn't matter 'where' that export comes from). I believe this can be configured in the Tesla Gateway which connects between the meter and consumer unit, directly from the meter tails (also acts as whole home backup in event of a power cut).

The only downside of this configuration is we will still experience modest clipping on our existing 3.645kWp array connected to a 3.0kW inverter, but we get 5kW of export so we can live with that.

So back to the installer to ask them to reapply under this fast track mechanism.

Martyn1981

I'm glad there's both a logical explanation, and a logical solution.

Just pondering out loud, but a 3kW inverter may be slightly more efficient for a 3.645kWp array (18% undersizing), so you might (just might) get back any losses from clipping that'll occurr on a cold sunny day.

NedS

Martyn1981 said:

Just pondering out loud, but a 3kW inverter may be slightly more efficient for a 3.645kWp array (18% undersizing), so you might (just might) get back any losses from clipping that'll occurr on a cold sunny day.

Hi Martyn,

Not sure I understand? Also, because the array is unshaded south facing, we do see near perfect profile and see clipping for much of the year. Looking back through my data from this year, I can see clear evidence of clipping in mid march whereas a perfect sunny day at the end of Feb showed no clipping. Likewise we clip all summer even with the higher temps. I suspect this continues until around mid September so we experience clipping for 6 months of the year, losing maybe ~2kW per day of generation at the height. Still, this is probably only costing us around £15/year at a SEG rate of 15p so not worth upgrading/replacing the 3kW inverter with a 3.68kW model.

Regarding your efficiency comment, given the Tesla PW3 will now only have 3.56kWp array attached, would it be better to configure the PW3 inverter to 3.68kW or 5kW? We are unlikely to be able to add more panels later so would matching the inverter setting to the size of the solar array (3.68kW) be preferable.

Martyn1981

Hi Ned, sorry, should have explained better. In the UK it's typically better to undersize an inverter by 10% to 20%, as we don't get the best sunny weather, but do 'enjoy' a lot of lower gen / cloudy weather. So it's about balancing some clipping against year round performance.

Given that hot panels will only generate about 80% to 90% of their rating, then some undersizing is preferable, as it typically means that the smaller inverter will be a bit more efficient in poorer generation, and should hopefully switch on a bit earlier, and switch off a bit later. Some studies* have shown that appropriately sized smaller inverters will generate more over the whole year, than an inverter that matches the array size.

It will also depend on location and orientation in the UK. For instance an E or W array will benefit more from a bit of undersizing, as they will see less direct sun, whereas a south facing array on the Cornish coast, that gets a lot of wind/breeze cooling may be better sized closer to 100% of array rating.

This doesn't apply to sunnier countries, where the array is expected to see strong sunlight for much of the year.

*I'm thinking back (and struggling to remember) one trial that was by SMA, possibly in the early 2010's. I believe they set up three 5kWp arrays, and compared a 3600TL inverter, to a 3800TL and a 4000TL. I recall that the highest generation for the year was the 3800, followed by the 3600, with the 4000 in last place.


With regard to the PW3, I'm way, way, way out of my comfort zone now, but I assume (hopefully someone can correct me) that it won't matter. The undersizing benefits are based on the physical components of the various inverters. So I assume that setting the same inverter to different limits won't make any difference, since you aren't changing it physically, only the rules by which it is to follow. Does that make any sense?

Qyburn

NedS said:

Regarding your efficiency comment, given the Tesla PW3 will now only have 3.56kWp array attached, would it be better to configure the PW3 inverter to 3.68kW or 5kW? We are unlikely to be able to add more panels later so would matching the inverter setting to the size of the solar array (3.68kW) be preferable.

Limiting PW3 output will also limit power that can come from the battery. So if set to 3.0kW then any time house load is over 3.0kW the extra will be drawn from the grid, even if the battery's full. 

Would you not be better putting all your solar, new and existing, into the PW3?

Nick_Dr1

Martyn1981 said:

Hi Ned, sorry, should have explained better. In the UK it's typically better to undersize an inverter by 10% to 20%, as we don't get the best sunny weather, but do 'enjoy' a lot of lower gen / cloudy weather. So it's about balancing some clipping against year round performance.

Given that hot panels will only generate about 80% to 90% of their rating, then some undersizing is preferable, as it typically means that the smaller inverter will be a bit more efficient in poorer generation, and should hopefully switch on a bit earlier, and switch off a bit later. Some studies* have shown that appropriately sized smaller inverters will generate more over the whole year, than an inverter that matches the array size.

It will also depend on location and orientation in the UK. For instance an E or W array will benefit more from a bit of undersizing, as they will see less direct sun, whereas a south facing array on the Cornish coast, that gets a lot of wind/breeze cooling may be better sized closer to 100% of array rating.

This doesn't apply to sunnier countries, where the array is expected to see strong sunlight for much of the year.

*I'm thinking back (and struggling to remember) one trial that was by SMA, possibly in the early 2010's. I believe they set up three 5kWp arrays, and compared a 3600TL inverter, to a 3800TL and a 4000TL. I recall that the highest generation for the year was the 3800, followed by the 3600, with the 4000 in last place.


With regard to the PW3, I'm way, way, way out of my comfort zone now, but I assume (hopefully someone can correct me) that it won't matter. The undersizing benefits are based on the physical components of the various inverters. So I assume that setting the same inverter to different limits won't make any difference, since you aren't changing it physically, only the rules by which it is to follow. Does that make any sense?

I'm struggling to understand the logic here. The inference is that bigger inverters are less efficient. This might be the case, but there is no real reason for this other than poor design. Indeed, normally for any power engineering exercise, the higher the rated design power, the more efficient things tend to be because even a fraction of a percent increase in efficiency leads to a lot less waste in the system, which means cooling is less of an issue. Its temperature that kills electronics eventually.
The only thing I can think of is the minimum string voltage spec for the inverter needs to be met so a small inverter may have a lower turn on voltage, conceivably allowing generation at lower light levels on a high peak voltage string, but for a lot of the time the inverter will be clipping as the max power for the inverter is reached. At this point the inverter is dissipating at its maximum value, which is the worst case for heat generation inside the inverter. The MPPT will be operating inefficiently as well.
Maybe this is why inverters don't last as long as maybe they should!

Happy to have it explained to me though.

Nick_Dr1

So here's a study from 2021 that compares two inverters (both rated at 3kW) over a years operation. You can read the details for yourselves, but the undersized inverter loses on all measurements. Here's the last line from the conclusion:-
"Thus, PV systems with undersized inverters will be losing electricity generation, in addition to reducing their useful life due to the stress of the components due to overtemperature, resulting in inverter changes over the life of the PV system. "
Overirradiance effect on the electrical performance of photovoltaic systems of different inverter sizing factors - ScienceDirect

Solar Energy

Volume 225, 1 September 2021, Pages 561-568 if the link doesn't work and you want to look it up manually.

NedS

Qyburn said:

NedS said:

Regarding your efficiency comment, given the Tesla PW3 will now only have 3.56kWp array attached, would it be better to configure the PW3 inverter to 3.68kW or 5kW? We are unlikely to be able to add more panels later so would matching the inverter setting to the size of the solar array (3.68kW) be preferable.

Limiting PW3 output will also limit power that can come from the battery. So if set to 3.0kW then any time house load is over 3.0kW the extra will be drawn from the grid, even if the battery's full. 

That was my concern too, but that is not what my installer and Tesla support have advised me when I phoned them.

The PW3 inverter is 11.04kW and is capable of supplying sustained loads of 11.04kW. When the inverter setting is limited in software to a lower value (either 5kW or 3.68kW in my case), presumably they are saying that this limit only affects/applies to the MPPT generation capacity, and does not affect what the battery can deliver under load. So, for example, if limited to 5kW and the house is pulling 10kW, a max of 5kW can come from solar and the remaining 5kW from stored battery capacity, but the total 10kW load can be met without having to draw from the grid.

If this is not the case, I will clearly be disappointed, as there are occasions where 5kW will be limiting (e.g, in winter the ASHP will be drawing maybe 1.3kW, the oven is on at 2.7kW to cook dinner, and the 3kW kettle is turned on giving a load of 7kW in addition to out ~300-500W background load)

Happy to hear from Tesla PW3 users as to their real world experiences with this.

Qyburn said:

Would you not be better putting all your solar, new and existing, into the PW3?

This is a good question and indeed is an option.

If I put all the solar into the PW3, I will have 7.2kWp solar connected to a 5kW inverter (that's the DNO limit of the offer), so I'll be clipping 2.2kW in summer (I currently see clipping on my 3.6kWp array on a 3kW inverter for ~6 months of the year - we are south facing, unshaded in a cooler area of the south of the UK). By the time we start clipping, the battery is already going to be full, so I'm giving up 2.2kWp of generation during the middle of the day on sunny days - for what benefit?

The original plan was to connect all the solar to the PW3, but the DNO would only permit 1kW of export if we did that with an inverter rating of 8kW.

So the best compromise seems to be to keep the original 3.6kWp of solar on the old inverter and add the newly installed 3.6kWp of solar on the PW3 (allowing full generation with no clipping), with an export limit of 5kW.

I will have enough solar connected to the PW3 in summer to fully charge it, so anything the old array/inverter generates is straight for export, so there are no/limited inefficiencies of multiple DC/AC/DC/AC conversions.

If I've missed some argument/reason why putting all the solar on the PW3 would be better, as I'm not understanding that at the moment?

Keeping both inverters, I can generate 3+3.6kW max (6.6kW) and can export 5kW, so providing I can use/store 1.6kW I'm getting as much out of the system as I can.

If I only use the PW3 inverter with all solar connected (using two MPPT strings for the two arrays), I can generate up to 7.2kW but I'm limited by the max 5kW inverter setting, but could potentially store the 2.2kW of clipping if there is spare capacity in the DC coupled battery, which is unlikely as the battery is already going to be full by mid morning on a sunny day.

Alnat1

I have a 3.6kW inverter and there's a setting "charge last" that sends the excess power to the DC battery only if generation is above 3.6kW. I set this when I know it's likely to be sunny most of the day and the "capped" amount will be sufficient for our evening and overnight use.

Is it possible you have similar settings available for your DC coupled battery?

Martyn1981

Nick_Dr1 said:

Martyn1981 said:

Hi Ned, sorry, should have explained better. In the UK it's typically better to undersize an inverter by 10% to 20%, as we don't get the best sunny weather, but do 'enjoy' a lot of lower gen / cloudy weather. So it's about balancing some clipping against year round performance.

Given that hot panels will only generate about 80% to 90% of their rating, then some undersizing is preferable, as it typically means that the smaller inverter will be a bit more efficient in poorer generation, and should hopefully switch on a bit earlier, and switch off a bit later. Some studies* have shown that appropriately sized smaller inverters will generate more over the whole year, than an inverter that matches the array size.

It will also depend on location and orientation in the UK. For instance an E or W array will benefit more from a bit of undersizing, as they will see less direct sun, whereas a south facing array on the Cornish coast, that gets a lot of wind/breeze cooling may be better sized closer to 100% of array rating.

This doesn't apply to sunnier countries, where the array is expected to see strong sunlight for much of the year.

*I'm thinking back (and struggling to remember) one trial that was by SMA, possibly in the early 2010's. I believe they set up three 5kWp arrays, and compared a 3600TL inverter, to a 3800TL and a 4000TL. I recall that the highest generation for the year was the 3800, followed by the 3600, with the 4000 in last place.


With regard to the PW3, I'm way, way, way out of my comfort zone now, but I assume (hopefully someone can correct me) that it won't matter. The undersizing benefits are based on the physical components of the various inverters. So I assume that setting the same inverter to different limits won't make any difference, since you aren't changing it physically, only the rules by which it is to follow. Does that make any sense?

I'm struggling to understand the logic here. The inference is that bigger inverters are less efficient. This might be the case, but there is no real reason for this other than poor design. Indeed, normally for any power engineering exercise, the higher the rated design power, the more efficient things tend to be because even a fraction of a percent increase in efficiency leads to a lot less waste in the system, which means cooling is less of an issue. Its temperature that kills electronics eventually.
The only thing I can think of is the minimum string voltage spec for the inverter needs to be met so a small inverter may have a lower turn on voltage, conceivably allowing generation at lower light levels on a high peak voltage string, but for a lot of the time the inverter will be clipping as the max power for the inverter is reached. At this point the inverter is dissipating at its maximum value, which is the worst case for heat generation inside the inverter. The MPPT will be operating inefficiently as well.
Maybe this is why inverters don't last as long as maybe they should!

Happy to have it explained to me though.

Yes, it's the lower power needed to operate it that improves the efficiency. hence why in countries like the UK with a less sunny climate, it's recommended to undersize the inverter 10% to 20% to maximise annual generation. That way the inverter will be running slightly more efficiently at all times generation is at or below the inverter limit, with a trade off of some clipping.

But since PV power drops off as the panels get hot, they rarely operate at 100% of rating, and even then, only for a few seconds or minutes. [But in clear April and October skies, with cold temps, they can max out for a bit longer.]

Edit - But just to repeat, for sunnier climes, it's not worth the trade off, as losses from clipping would be greater than the increased efficiency of a smaller inverter, and said inverter would be working hard(er) for longer. Whereas a UK inverter gets a pretty gentle life.